The present invention relates to a method for the production and use of hydrogen donor solvents to increase heavy oil-to-hydrocarbon distillate conversion efficiency.
Terminology is important, especially for a complex field like hydrocarbon processing that progressed in parallel and very non-linear scientific and engineering practice pathways. Originally, heavy oils were hydrocarbons with a high density for a given boiling point range. However, the term heavy oil is often used interchangeably with high boiling by practicing engineers because most oil fractions with higher densities also have higher boiling points. However, some highly paraffinic oils or oil fractions may have significantly higher boiling points than much heavier, i.e., denser, aromatic oils or oil fractions. For the purposes of this invention, a heavy oil contains a significant quantity of a high density vacuum residual oil. Residual oils, also called residua or resids, are typically those fractions which are non-distillable under given conditions and remain at the bottom of a vacuum distillation tower and have equivalent normal boiling point (NBP) greater than approximately 525° C.
The efficiency of processes to convert heavy oils to distillates is generally determined by the relative rates of cracking reactions to produce lower molecular weight species and the rate of free radical polymerization reactions to produce higher molecular weight and less soluble species. The polymerization reaction rate dramatically accelerates to form solid petroleum coke when the polymerization reaction products form a separate mesophase. Therefore, control of both the heavy oil conversion and process solvent and solute properties are important.
The solubility of residual oil components in alkanes (paraffins), e.g., propane, butane, pentane, hexane, and heptane) has been used by petroleum refiners to up-grade residual oils and by researchers to obtain more detailed information about these component properties. A two-product commercial deasphalting unit produces deasphalted oil (DAO) and asphaltene streams and a three-product commercial deasphalting unit produces DAO, resin, and asphalt streams. The DAO, resin, and asphalt stream properties vary over a wide range depending on the deasphalter operating conditions. Broadly, the deasphalter product aromaticity and molecular weight have the following ranking: DAO<resin<asphaltene. Petroleum chemists use similar terms, with substantially different meanings, to specify residual oil solubility classes. For petroleum chemists, asphaltenes and maltenes are terms used to describe the insoluble and soluble fractions of a vacuum residue or deasphalter asphalt product. They are defined by the respective insolubility and solubility of these fractions in light hydrocarbons such as n-pentane, n-hexane, or n-heptane. As a result, pentane-insoluble-asphaltenes would have a lower molecular weight and aromaticity than heptane-insoluble-asphaltenes. The petroleum chemists usually define oils and resins as maltene species that readily adsorb on a packing and can be readily desorbed using alkane and polar solvents, respectively.
This invention defines (1) coke precursors as marginally soluble species in the heavy oil conversion process solvent and (2) process solvent properties in terms of the equivalent deasphalter DAO and asphaltene residual product properties. Theoretical methods to estimate solubility can be used to analyze heavy oil conversion process data [Jianzhong Wu, John M. Prausnitz, and Abbas Firoozabadi, “Molecular-Thermodynamic Framework for Asphaltene-Oil Equilibria”, AlChE Journal, Vol. 44, No. 5, May 1998].
This invention provides a heavy oil-to-distillates conversion method that is differentiated from and superior to the related art. This task is complicated by the fact that both this invention and the related art utilize very complex and poorly understood thermal cracking reactions to convert heavy oil to distillates. Molecular weight, elemental analysis, nuclear magnetic resonance (NMR) spectroscopy, and X-ray diffraction (XRD) analyses can be used to estimate the average structural data for hydroconversion feed and products [George Michael, Mohammad Al-Siri, Zahida Hameed Khan, and Fatima A. Ali, “Differences in Average Chemical Structures of Asphaltene Fractions Separated from Feed and Product Oils of a Mild Thermal Processing Reaction,” Energy Fuels, Vol. 19, No. 4, pages 1598-1605, 2005]. Even these very time consuming and expensive analytical methods provide only very general guidance to assess process performance. As a result, heavy oil process developers are forced to use less rigorous and costly methods to characterize, evaluate, and improve heavy oil conversion processes. Process development teams tend to use somewhat different approaches to analyze their processes and assess their performance relative to alternative approaches. The present inventor has discovered that the very general reaction system on FIG. 1 provides a useful framework to assess and guide the development of the present invention. Broadly, this oversimplified process framework envisions that heavy oils are converted to distillates via thermal cracking reactions, which also initiate free radical polymerization reactions that are terminated by either hydrogen transfer or coking reaction.
More specifically, the conversion process is initiated by thermal cracking of a carbon-carbon bonds (R—R′, where R and R′ represent the feedstock structure on either side of the ruptured bond) via Reaction 1 to form short lived free radical intermediate species (R. and R′.). These unstable free radical species can react with labile hydrogen atoms in the heavy oil to produce the desired stable reaction product via Reaction 2 (or Reaction 6). The labile hydrogen is typically a naphthenic hydrogen atom that is bonded to a carbon atom that is in the alpha position relative to an aromatic carbon (see hydrogen donor diluent example on FIG. 1). Hydrogen donor diluents are generally highly aromatic distillates, e.g. fluid catalytic cracking cycle (decant) oils, thermal tars, or coker gas oils. The hydrogen donor diluent can be regenerated via Reaction 5. Residual oil species, particularly the highly aromatic asphaltene and resin components, can provide labile hydrogen to produce stable cracked products via Reaction 6 using an unsupported hydrotreating catalyst. Since asphaltene species are generally too large to be effectively hydrogenated using a support hydrotreating catalyst, asphaltene fraction labile hydrogen species are most effectively produced using an unsupported colloidal catalyst hydrotreating catalyst via Reaction 7. Free radical polymerization reactions (Reaction 3) can produce progressively larger and less soluble species (R—R′—R. & R—R′—R′.) until these species reach the solution solubility limit, form a separate mesophase, and then very rapidly produce the less desirable solid coke product via Reaction 4. This framework will be used in the discussion of the related art.